This is a tricky question as there are many statistics done on the subject with contradictory results, probably because the concept of failed differs from one company to the other. For example, some companies think that projects that are behind time and over budget are failed projects, while others think that these projects are challenged, and failed projects are those who are canceled or killed during the timeline of the project because of one of the following reasons:
- The project is no longer expected to give the company the expected ROI
- The project now conflicts with the company's business objectives
- The company can no longer fund the project
I think the above 3 reasons can include any other reason for projects to be canceled or killed.
In the software industry, failure rate, according to the Standish Chaos report, is around 30%, but if you include the "challenged" rate, then the failure rate will be over 70%. That's a lot. Of course, this report is rebuked and refused by many leaders in the project management profession.
In the construction industry, the failure rate is much lower, because of the maturity of the field. Many projects in the construction industry go over budget and get delayed, but, in the construction world, that's rarely considered a failure. Although there are no formal statistics on the failure rate in construction project management, I believe the number is below 10%.
As for the aviation industry, because it is very small (you only have a handful of companies producing planes), the failure rate can be highly skewed (look at the A380 project for example) and thus no reliable statistics can be given in that area.